ERP News


Why Enterprises Need To Focus On Decentralized AI

591 0
Decentralized AI

Decentralized AI

Decentralized AI- 2020: Year of Worldwide Data Crisis?

In 2009 when we (me and Pavan Vara, a brilliant scientist and researcher from IIT, Madras) wrote our paper on cloud computing we worried deeply about the state of the cloud computing industry from a global energy perspective. So we proposed an extensible architecture in which we propose to connect the APIs (Power/MDM APIs, Data Center APIs and Consumption-based Open Standards API, etc.) through and through the whole value chain.

Out of our concern for the energy crisis which we foresaw back then, I wrote:

With more and more Data Centers being built, and soon after we are done bashing the economy and again ready to make hay as the economy starts turning around, we can alter the shape of our destiny dramatically!

As other industries will sag, think of hotels, airlines, etc., which will certainly contribute to a certain reduction of carbon emissions, there is strong reason to believe that the spin-offs from the remote-everything will go into overdrive.

Today we are carefully talking about online conferences and soon there will be lots of online activities that will be firing up all over the place.

This will lead to huge data crunch operations as more and more information will need more processing power as it will come in audio, video and other formats.

This will be extremely demanding for data centers, no matter how centralized they are. The data center owners will have to realize that their continuous dependence on energy suppliers and other crises shouldn’t lead to a financial market-like collapse.

There is again strong reason to believe that the inter-dependence will also foster co-creation and co-conservation but it could also lead to frictions when it’s time for “someone to pay.”

We had focused our paper on exactly addressing the same problem. We tried to combine the “industrialized metering solutions”; something the electrical grids have come to do in a reasonable and sustainable fashion, while still making enough profit.

The need to standardize and meter the data centers and eventually charge the consumer with an RTP/RTB (Real-Time Pricing/Real-Time Billing) mechanism which is state-approved — for the same reason I keep calling it RCC or Regulated Cloud Computing.

Hmmm, we proposed regulated cloud computing and I am not really sure if that happened but one thing did happen for sure!

A lot of data was created and data-hungry companies like Facebook, Google and Twitter were growing like weeds!

Read More Here

Article Credit: Forbes

Leave A Reply

Your email address will not be published.